EP2469366B1

EP2469366B1 - Controlling JIT item production via kanban cards

Info

Publication number: EP2469366B1
Application number: EP10197022.6A
Authority: EP
Inventors: Paolo Fontanot; Antje Haase; Fabio Sala
Original assignee: Siemens AG
Current assignee: Siemens AG
Priority date: 2010-12-27
Filing date: 2010-12-27
Publication date: 2013-08-14
Anticipated expiration: 2030-12-27
Also published as: US20120165968A1; CN102591268B; EP2469366A1; US8805564B2; CN102591268A

Description

The present invention relates to a method and to a system for controlling JIT item production via kanban cards according to the preamble of claims 1 and 8 respectively.
In the manufacturing field, with the term "flow shop" it is indicated a discrete manufacturing production system where items are transformed and assembled according to well defined routes that follow the process layout.
In the flow shop, it is possible to identify one or more sequences of production stages where the items are manufactured. Such production stages are also denoted also as workstations. Moreover, at the flow shop, the process consists of a series of production steps and the operations of every productive job are processed on machines and workstations in the same order.
In few words, in the shop flow, the work shall follow a fixed path.
Typically, in the flow shop, between the different workstations there is a buffer for work in process inventories in order to decouple the process of each workstation and to safeguard against random fluctuations of the production pace as well as of the demand for items.
In several cases, the production flow in the flow shop can be controlled using Just-In-Time (JIT) methods like the "kanban" method, with the goal of synchronizing the production achieving at the same time the maximum service level, i.e. the capability to satisfy the product demand, and the lowest possible amount of waste due to inefficiencies and work in process inventories.
The kanban method consists in visual signals/tickets/cards, called "kanbans" or "kanban cards", that are used to authorize production activities without the need for releasing and dispatching production orders to the shop floor. Workstations are authorized to operate only when a production kanban card is received: a fixed amount of material (either a subassembly or a final product) is produced for each kanban card and eventually moved to the downstream stage when another movement kanban card is issued to authorize the movement.
Kanban cards tell the operators when to start and stop producing at a production stage and when to move the items from a production stage to the next one.
Figure 1 is a diagram schematically illustrating a flow shop type of production system using a kanban method. In Figure 1, the customer demand CD for production of some finished good material FG is originated by customers C1, C2. At the beginning of the production flow, a raw material buffer B_RM comprises raw material for the production of items for work in process material M. The production flow comprises N production pairs, each comprising a workstation W_j and its assigned buffer B_j. The buffer B_N of the last production pair contains the produced finished good items FG intended to be sent to the customers C1, C2. The first stage of the production flow is performed at the first workstation WS₁ and the last stage of the production flow is performed at the last workstation WS_N. Between the first workstation WS1 and the last workstation WS_N there are one or more intermediate workstations WS_j. Information on items to be produced is conveyed via kanban cards K between each workstation W_j. It is noted that the stream of information flow is the opposite than the stream of material production flow, i.e. downstream workstations with respect to one given workstation are located at its right while upstream workstations with respect of a given workstation are located at its left.
Kanban methods may use cards, boards, lights, empty containers or any other means to signal the need for an item. Each kanban card corresponds to a fixed amount (one or more) of items to be produced and moved at the same time. The mechanism of kanbans regulates the production flow between all the stages of the production process. The buffer stock level in every buffer between the different production stages is limited by the number of kanban cards available in the process.
For each specific item, the available quantity of material can be roughly measured on the basis of the number of "full" kanban cards for that item, i.e. the number of kanban cards between two production stages assigned to a particular item or part number that are in the "full" status.
In fact, kanban cards are typically assigned with one of the following two statuses:

1. an "empty status": when the material piece is used by the downstream stage, the kanban status is updated so that a signal is sent to the previous stage for material replenishment;
2. "full status": when the material piece is produced, the kanban status is updated and the container (material and kanban card) is moved to the buffer (available for the downstream stage).

Advantageously, by limiting the number of kanban cards also the waste due to inventory is limited.
Kanban methods have optimum performances in scenarios in which the demand is relatively constant.
Unfortunately, in some other scenarios, kanban methods may present some drawbacks.
Examples of problematic scenarios are flow shops in which multiple items are produced at the same time, where at some stages of the production process, it is possible that more than one kanban card per item arrive simultaneously, requiring the production of different items of different type at the same time.
However, since, usually, in a workstation only one item at a time can be manufactured, the workstation operator has to select which is the kanban card of a specific item type to be processed first. This operator's choice may be free or, most probably, follows one of the pre-defined methods of the specific company, e.g. First-In-First-Out (FIFO).
It is evident how this choice may have problematic impacts. In fact, this choice is performed by each operator at a local level, i.e. at the level of the workstations, but this choice affects the overall performances of the whole production process. In fact, the workstation operators, who are involved in a kanban production flow, have typically no possibility to foresee a variation in demand by the end-customer because their visibility is limited to the incoming kanban cards at their own workstation. These kanban cards only tell the operator to produce new items to replace the ones consumed by the immediately next production step or by the end customer.
In case different kanban cards for different item types arrive to a workstation from more than one consuming resource, the operators are typically not prepared to react correctly so as to minimize the drawbacks caused by wastes from overproduction or by the delays from underproduction.
Moreover, real world operational conditions amplify these problems because the theoretical hypotheses for using kanban, i.e. relatively constant demand, are often not met. In general, kanban methods have best performances with a relatively smooth rate of the demand, a "constant" production lead time, a short set up time and repetition of demand. In reality, the demand rate is rarely constant and, instead, low and unpredictable fluctuations on demand have to be managed. Sometimes there are bigger fluctuations due to the request for new items not already present in the production flow (phase-in of new products) or due to stopping producing some other items (phase-our of a product).
Increases or decreases of the demand, even if they are little, have an impact on the production flow and on the material supply. If some process varies its withdrawal of items, the range of these fluctuations will increase as they move up the line towards the earlier processes. This is known as demand amplification. The variation of product demand acts like a wave through the entire production process and impacts at staggered time all production steps. For example, this means that, when all work stations are busy producing some item, the effect of a demand fluctuation at the last workstation WS_N propagates slowly to the upstream work stations and reaches the first workstation WS₁ after N times the average throughput time of a work station. The effect of this slow propagation of the demand "wave" is a global delay of the productive system in responding to fluctuations of the demand for finished product FG.
To better understand this concept, let us consider the following limit situation: when an item of a specific type is not required anymore, the operator at the workstation does not detect this change immediately and may continue to produce the material M that is not required anymore because he/she still have some empty kanban cards to be fulfilled. With production terminology, one can say that the workstation operator may be producing waste when the demand decrease and he/she may be producing in delay when the demand rises Prioritization of different kanban cards requiring different item types to be produced has been explored and utilized in different situations. The downstream adjacent workstation which generates an empty kanban card can notify the urgency of the underlying message by using for example different colors for the issued kanban card, like for example with the Red-Amber-Green (RAG) method applied to individual kanban cards.
According to the RAG kanban method, it is possible using a traffic light RAG system to prioritize the work of the supplying workstation, since it may receive replenishment signals from more than one consuming resource simultaneously. The traffic light color-coding system is using three colors: Red means needs "urgent action" (out of control, shortage imminent), Amber means "going out of control" (on the borderline, needs reordering) and Green means "no problem" (within acceptable limits).
In the following example, the RAG method applied to individual kanban cards is illustrated. Let us assume we have a molding shop at a motorbike manufacturer supplying eleven different assembly lines with a variety of components. Let us assume that there are constant assembly shortages in some assembly lines while other assembly workstations are climbing over moldings. In order to improve such problematic situation, the total kanban population for each item may be equally divided into three color-coded RAG groups. For example, green kanbans are to be used first until they are exhausted, followed by amber kanbans and then red kanbans. Advantageously, by using this RAG kanban method, the work-in-process control problems may disappear since the system operates very happily with green and amber kanbans setting priorities. Instead, the appearance of a red kanban in the molding shop causes significant attention to be paid to it.
The RAG kanban method applied to individual kanban cards has the advantage of simplicity but the disadvantage of being not much flexible.
In fact, with this method, once a kanban card has been issued with a given priority, this priority remains unchanged even if there is a change on the demand for products.
In their most simple and traditional forms, kanban methods rely only on the exchange of visual signals like cards or empty containers.
Known improvements of traditional kanban methods are Electronic Kanban (eKanban) techniques which provide with more flexibility in improving the Kanban process flow. The eKanban systems comprise all the basic components of traditional Kanban systems whilst using IT systems also in order to replace traditional kanban cards with barcodes and electronic messages. The eKanban system is a signaling system that can use a mixture of different technologies to automatically trigger in real time material movement and/or production activities at the workstations. The data transfer can be done either manually by an operator entering the data or automatically via a barcode reader, RFID chips or other methods so as to reduce information delivery time. Kanban cards can be replaced in eKanban by "virtual Kanban cards" displayed on a computer screen on a virtual Kanban board. The eKanban signal notifies the need of restocking of items when a predefined minimum of stock level is reached at each local buffer.
The RAG kanban and eKanban systems solve the problem of dealing with simultaneous requests but not the problem of containing the impact of demand propagation fluctuations.
It is therefore the aim of the present invention to overcome the above mentioned drawbacks, for controlling JIT item production via kanbans in a discrete manufacturing production system which minimize the impact of demand propagation fluctuations.
The aforementioned aim is achieved by a method and a system for controlling JIT item production via kanban cards in a discrete manufacturing production system, wherein items are transformed according to a flow shop comprising a raw material buffer followed by a sequence of production pairs comprising one workstation and its assigned buffer; wherein each workstation produces an item to be moved into its assigned buffer; wherein the first workstation produces an item starting from the raw material buffer and each other remaining workstation produces an item starting from the buffer assigned to the previous adjacent workstation; wherein each workstation is able to produce more than one type of item according to requests received by means of different sets of different kanbans (K) associated to different item types; wherein each set of kanbans consists of a given number of kanbans;
wherein each kanban can have two statuses:

an empty status denoting that item replenishment is requested and
a full status denoting that the produced item is ready in the buffer;
the invention comprising the following:
1. a) for each production pair, for each specific item type and at given points in time, providing a local buffer threshold parameter denoting the desired safety stock of the items of the specific type in the buffer;
2. b) for each production pair, for each specific item type and at other given points in time, providing a global demand threshold parameter denoting the quantity of items of the specific item type required along the downstream process to satisfy the estimated demand;
3. c) at each production pair, whenever a kanban associated to an item of a specific type switches status, assigning to it one of four priority levels indicating the priority level of the production request of the associated item type, according to the following logical rules:
  1. I) a first priority level is assigned to the kanban if the number of full kanbans of items of the same specific type at the production pair buffer is greater than the local buffer threshold and if the number of full kanbans for the items of the same specific type along the downstream process is greater than the global demand threshold;
  2. II) a second priority level is assigned to the kanban if the number of full kanbans of items of the same specific type at the production pair is lower or equal than the local buffer threshold and if the number of full kanbans for the items of the same specific type in the downstream process is greater than the global demand threshold;
  3. III) a third priority level is assigned to the kanban if the number of full kanbans of items of the same specific type at the production pair is greater than the local buffer threshold and if the number of full kanbans for the items of the same specific type along the downstream process is lower or equal than the global demand threshold;
  4. IV) a fourth priority level is assigned to the kanban if the number of full kanbans of items of the same specific type at the production pair is lower or equal than the buffer threshold and if the number of full kanbans for the items of the same specific type along the downstream process is lower or equal than the demand threshold;
4. d) at each production pair, producing an item of a type whose priority level is the highest according to a given priority model defining the priority sequence of the four priority levels for the production requests for different item types.

In invention embodiment, in item c) with the sentence "whenever a kanban associated to an item of a specific type switches status" it may preferably be meant:

"whenever a kanban associated to an item of a specific type switches status from empty to full or from full to empty" or
"whenever a kanban associated to an item of a specific type switches status from empty to full or whenever a kanban associated to an item of a specific type switches status from full to empty".

In invention embodiments, referring as production pair "j" the generic production pair of the "N" production pairs where "j" is comprised between "1" and "N"; and wherein at step b) the global demand threshold parameter for production pair "j" may preferably ne calculated as the number of finished good items of the specific type, required in the average lead time interval, defined as the average time for an item of a certain type to move from buffer "j-1" to buffer "N" of the production flow; and
wherein the number of full kanbans for the items of the same specific type along the downstream process may be calculated as the sum of the full kanbans comprised in each of the buffer belonging to each production pair comprised between production pair "j-1" and production pair "N".
In invention embodiments, kanban cards may be conveniently assigned to different priority levels are marked with different colors.
In invention embodiments, kanban may advantageously be marked with colors as follows:

a kanban of the first priority level is assigned with a green color;
a kanban of the second priority level is assigned with a blue color;
a kanban of the third priority level is assigned with a yellow color;
a kanban of the forth priority level is assigned with a red color.

In invention embodiments, wherein information on kanban cards may conveniently be visualized on a screen via an electronic kanban-board at the workstation via a Manufacturing Execution System which is provided with the real time information on the movements of produced items and of the kanban cards.
Furthermore, a computer program element can be provided, comprising computer program code for performing steps according to the above mentioned method when loaded in a digital processor of a computing device.
Additionally, a computer program product stored on a computer usable medium can be provided, comprising computer readable program code for causing a computing device to perform the mentioned method.
The proposed invention enables to timely react to changes in the customer demand for finished goods. Thus, the invention embodiments minimize the negative impacts of fluctuations in the finished good demand which may produce slow wave propagation through the production stages.
With embodiments of the proposed invention, the prioritization of production of items of different types is signaled via a priority model which may provide visual and/or intuitive instructions to workstation operator.
Embodiments of the proposed invention are compatible with the JIT philosophy and enhance the JIT capabilities of a production flow.
The invention will now be described in preferred but not exclusive embodiments with reference to the accompanying drawings, wherein:

Figure 1: diagram which schematically illustrates a flow shop using a Kanban method (Prior Art, already described);
Figure 2: diagram which schematically illustrates a flow shop in accordance with an example embodiment of the present invention.

In the drawings, like reference signs refer to same or similar elements.
According to the proposed invention, the item manufacturing path is organized in a flow shop with JIT control and kanban cards.
Figure 2 is a diagram schematically illustrating a production flow in accordance with an example embodiment of the present invention.
As shown in Figure 2, items M are transformed according to a production flow comprising a raw material buffer B_RM followed by a sequence of production pairs W₁,B₁; W_j,B_j; W_N,B_N comprising one workstation W_j and its assigned buffer B_j. Each workstation W_j produces, by consuming items M comprised in the adjacent previous buffer B_j-1, an item M to be moved into its assigned buffer B_j. The first workstation W₁ produces an item M starting from items comprised in the raw material buffer B_RM and each other remaining workstation produces an item starting from items comprised in the buffer assigned to the previous adjacent workstation. According to the present invention, each workstation W_j is able to produce more than one type of items according to requests received by means of different sets of different kanbans cards EKB associated to different item types. In the embodiments of Figure 2, it is present a MES system receiving real time information RIT from each buffer B_j and signaling production requests to each workstation W_j by means of electronic kanban cards EKB. The skilled in the art easily understands that electronic kanban cards may be physical or virtual cards.
For each type of item to be produced are provided a set of kanban cards consisting of a given number of cards. Each kanban card can have two statuses: 1) an empty status denoting that item replenishment is requested (since the item is being used by the following adjacent workstation) and 2) a full status denoting that the produced item is ready in the buffer.
For each production pair W₁,B₁; W_j,B_j; W_N,B_N, for each specific item type and at given points in time, a local buffer threshold parameter is provided for denoting what is the wished safety stock of the items of the specific type in the buffer of the production pair.
For each production pair W₁,B₁; W_j,B_j; W_N,B_N, for each specific item type and at other given points in time, a global demand threshold parameter is provided for denoting the quantity of items of the specific item type required along the downstream process to satisfy the estimated demand.
The value of the local and global threshold parameters are calculated at some given points in time which may be the same or may differ for the local and global thresholds.
The given points in time may be periodically set or may be triggered by some events as, for example, the change of any of the variables at stake, e.g. every time an item of the specific type is produced and/or every time an item of the specific type is consumed.
At each production pair W₁,B₁; W_j,B_j; W_N,B_N, whenever a kanban card associated to an item of a specific type switches status from full to empty or from empty to full (alternatively or in both situations), the kanban card is assigned to one of four priority levels indicating the priority level of the production request of the associated item type.
The priority level is assigned according to the following logical rules:

I) a first priority level is assigned to the kanban card if the number of full kanban cards of items of the same specific type at the production pair buffer is greater than the local buffer threshold and if the number of full kanbans for the items of the same specific type along the downstream process is greater than the global demand threshold;
II) a second priority level is assigned to the kanban card if the number of full kanban cards of items of the same specific type at the production pair is lower or equal than the local buffer threshold and if the number of full kanbans for the items of the same specific type in the downstream process is greater than the global demand threshold;
III) a third priority level is assigned to the kanban card if the number of full kanban cards of items of the same specific type at the production pair is greater than the local buffer threshold and if the number of full kanbans for the items of the same specific type along the downstream process is lower or equal than the global demand threshold;
IV) a fourth priority level is assigned to the kanban card if the number of full kanban cards of items of the same specific type at the production pair is lower or equal than the buffer threshold and if the number of full kanbans for the items of the same specific type along the downstream process is lower or equal than the demand threshold.

At each production pair W₁,B₁; W_j,B_j; W_N,B_N, it is produced an item of a type whose priority level is the highest. The highest priority level is defined according to a given priority model. The priority model provides the sequence of priorities for each of the four priority levels in order to fulfill the production requests for different item types. Thus, at each production pair the operator produces the item of a type at highest priority level according to the defined priority model.
Advantageously, the four priority levels combine information about local status and global status of the production request for items of a given type at each production pair. The local status provides an indication of the capability to satisfy the requirements of the adjacent following production stage.
Instead, the global status provides an indication of the capability to satisfy the customer demands for finished goods.
The local status is computed by comparing the buffer level, i.e. the amount of available items, at the given production stage pair, with a predefined local safety threshold representing a safety stock to compensate for local demand fluctuations.. When the buffer level locally goes below such threshold, this item becomes urgent.
The global status is computed by comparing the buffer level at all other production stages with the estimated demand for these items. Such estimated demand is strictly related to the demand of finished products, calculated by exploding the bill of materials and taking in account the average throughput time of each production stage.
In embodiments of the present invention, the global demand threshold parameter for the workstation in generic position "j" (where "j" can assume values between "1" and "N", both extremes included) may be advantageously calculated as the number of finished good items FG demanded by the customer CD in the time interval between "current time" and "current time + Tj"; where Tj is the average lead time of a piece of a certain type to move from buffer B_j-1 to buffer B_N (the last buffer in the production flow).
In embodiments of the present invention, the number of full kanban cards for the items of the same specific type along the downstream process may be calculated as the sum of the full kanban cards comprised in each of the buffer belonging to each production pair comprised between production pair "j-1" and production pair "N". Thus, the global demand is covered at the current production rate when the number of pieces (full kanban cards) contained in the buffers between j-1 an N is bigger then the global demand threshold parameter as defined above.
In embodiments of the present invention, different color signals may be assigned to the different kanban priority levels as follows.
A kanban assigned with the first priority level may be marked with a green color, indicating a "regular" status for the production of the specific item since both its local and its global statuses are not critical.
A kanban assigned with the first priority level may be marked with a blue color, indicating a "sub-critical" status for the production of the specific item since its local status is critical but its global status is not critical.
A kanban assigned with the first priority level may be marked with a yellow color, indicating a "expedite" status for the production of the specific item since its local status is not critical but its global status is critical.
A kanban assigned with the first priority level may be marked with a green color, indicating a "critical" status for the production of the specific item since both its local and its global statuses are critical.
In this embodiment, a visual, color coded signal is advantageously used to give to the workstation operators an immediate indication of the priority of each specific item. In other embodiments, the light color may be also switched off, in a fifth priority level in case there is no request for that specific item.
In a preferred embodiment of the invention, information on the kanban cards and priorities may be visualized on a screen via an electronic kanban-board EKB at the workstation via a Manufacturing Execution System MES as shown in Figure 2.
The electronic kanban board screens EKB may show, at each production stage/workstation, a color coded signal for each item to be produced, here denoted as part number, listing the electronic kanban cards sorted by item.
For example, the electronic kanban board EKB for the first workstation W1 may be represented by the following sample table, Table 1: Table 1: sample table

Part
Number Number of
Kanban cards Priority

xyz 3 Blue

xxx 10 Red

yyy

0

zxy 5 Yellow

Zzz 6 Green
The logic of the MES system computes in real time, via real time information RIT provided manually or automatically via different means, the priority level/color for each kanban card of each part number.
In the sample embodiment shown in Figure 2, a Manufacturing Execution System MES supervises the production flow with a full visibility on all items involved in the production process.
A logic implemented in the MES system computes the local and global statuses of material at each production stage, based on the available quantity on each buffer. Advantageously, these statuses are used to produce visual signals, e.g. colored lights, at each production stages that are based on the buffer level on the current production step while looking also at the downstream buffer levels to give indications to the workstation operators about the production priority for each specific item.
Conveniently, the negative impacts due to wave demand propagation are minimized. In fact, with invention embodiments, by supervising the entire production flow, the propagation of demand fluctuations through the production flow is anticipated since each production step is provided with information on what is happening on all the following production steps and not just on the adjacent one. Workstation operators are thus provided with the visibility which enables them to timely react by making the right choices, especially in case of demand changes they are able react according to the disturbances in the flow control. Instead of reacting after receiving an incoming "wave" of demand fluctuation, it is possible to anticipate it by choosing the items to produce first. In case of demand decrease for a specific item, workstation operators can quickly react by avoiding the waste related to the production of specific items not immediately requested. In case of demand decrease for a specific item, workstation operators can quickly react by anticipating the production of some items, avoiding receiving requests that cannot be satisfied on time.
In other invention embodiments, at the intermediate storage point or at the buffer for the items just manufactured at each workstation, the place for each different specific item, i.e. part number, may be equipped with a physical light signal and a holder to keep the kanban cards for that part number. Moreover, kanban cards may contain an inexpensive RFID tag and the storage point may be equipped with an RFID antenna. With RFID tags, the MES may automatically be provided with the real time information RIT on the movements of produced items and of the kanban cards.

List of Acronyms

IT: Information Technology
JIT: Just In Time
MES: Manufacturing Execution Systems
RAG: Red Amber Green
RFID: Radio Frequency Identification

Claims

A method for controlling JIT item production via kanban cards in a discrete manufacturing production system,
wherein items are transformed according to a flow shop comprising a raw material buffer (B_RM) followed by a sequence of production pairs (W₁, B₁; W_j,B_j; W_N, B_N) comprising one workstation (W_j) and its assigned buffer (B_j); wherein each workstation (W_j) produces an item (M) to be moved into its assigned buffer (B_j); wherein the first workstation (W₁)produces an item (M) starting from the raw material buffer (B_RM) and each other remaining workstation produces an item starting from the buffer assigned to the previous adjacent workstation;
wherein each workstation (W_j) is able to produce more than one type of item according to requests received by means of different sets of different kanbans (K) associated to different item types; wherein each set of kanbans consists of a given number of kanbans;
wherein each kanban can have two statuses:
- an empty status denoting that item replenishment is requested and

- a full status denoting that the produced item is ready in the buffer;
the method characterized in that it comprises the following steps:
a) for each production pair (W_j,B_j), for each specific item type and at given points in time, providing a local buffer threshold parameter denoting the desired safety stock of the items of the specific type in the buffer;

b) for each production pair (W_j,B_j), for each specific item type and at other given points in time, providing a global demand threshold parameter denoting the quantity of items of the specific item type required along the downstream process to satisfy the estimated demand;

c) at each production pair (W_j,B_j), whenever a kanban associated to an item of a specific type switches status, assigning to it one of four priority levels indicating the priority level of the production request of the associated item type, according to the following logical rules:
I) a first priority level is assigned to the kanban if the number of full kanbans of items of the same specific type at the production pair buffer is greater than the local buffer threshold and if the number of full kanbans for the items of the same specific type along the downstream process is greater than the global demand threshold;

II) a second priority level is assigned to the kanban if the number of full kanbans of items of the same specific type at the production pair is lower or equal than the local buffer threshold and if the number of full kanbans for the items of the same specific type in the downstream process is greater than the global demand threshold;

III) a third priority level is assigned to the kanban if the number of full kanbans of items of the same specific type at the production pair is greater than the local buffer threshold and if the number of full kanbans for the items of the same specific type along the downstream process is lower or equal than the global demand threshold;

IV) a fourth priority level is assigned to the kanban if the number of full kanbans of items of the same specific type at the production pair is lower or equal than the local buffer threshold and if the number of full kanbans for the items of the same specific type along the downstream process is lower or equal than the global demand threshold;

d) at each production pair, producing an item of a type whose priority level is the highest according to a given priority model defining the priority sequence of the four priority levels for the production requests for different item types.
The method according to claim 1, where at step c) with the sentence "whenever a kanban associated to an item of a specific type switches status" it is meant:
- "whenever a kanban associated to an item of a specific type switches status from empty to full or from full to empty" or

- "whenever a kanban associated to an item of a specific type switches status from empty to full" or

- "whenever a kanban associated to an item of a specific type switches status from full to empty".
The method according to any of the preceding claims, wherein referring as production pair "j" the generic production pair of the "N" production pairs where "j" is comprised between "1" and "N"; and wherein at step b) the global demand threshold parameter for production pair "j" is calculated as the number of finished good items of the specific type, required in the average lead time interval, defined as the average time for an item of a certain type to move from buffer "j-1" to buffer "N" of the production flow; and wherein the number of full kanbans for the items of the same specific type along the downstream process is calculated as the sum of the full kanbans comprised in each of the buffer belonging to each production pair comprised between production pair "j-1" and production pair "N".
The method according to any of the preceding claims, wherein kanbans assigned to different priority levels are marked with different colors.
The method according to claim 4, wherein kanban are marked with colors as follows:
- a kanban of the first priority level is assigned with a green color;

- a kanban of the second priority level is assigned with a blue color;

- a kanban of the third priority level is assigned with a yellow color;

- a kanban of the forth priority level is assigned with a red color.
The method according to any of the previous claims, wherein information on kanban cards are visualized on a screen via an electronic kanban-board (EKB) at the workstation via a Manufacturing Execution System (MES) which is provided with the real time information (RIT) on the movements of produced items and of the kanban cards.
The method according to any of the previous claims, wherein kanban cards comprise RFID tag and the buffers are equipped with an RFID antenna.
The method of any of the claims 1 to 7 characterized in that it is implemented in software.
A system having means for performing the steps of the method according to any of the claims 1 to 8.
A computer program product for performing steps of the method according to any of the claims 1 to 8.